Hi, my name is Nicolás Aller Ponte
Data Science & Engineering Student.

Know more

About me

Nicolás Aller Ponte

I'm a Data Science & Engineering student at Universidade da Coruña. I worked as Research Assistant at the Inditex-UDC AI Chair in Green Algorithms (until March 2026), where I developed feature-selection systems for AI models. I've also worked as an AI Engineer at Gradiant, building computer vision pipelines for car damage detection.

My stack spans the full data lifecycle — from HPC and parallel computing (C/C++, CUDA, MPI) to deep learning (PyTorch, TensorFlow), data engineering (Spark, SQL, Docker, AWS/GCP), and NLP/Computer Vision.

CV

Projects

Outfit Intelligence — HackUDC 2026 🏆

Double award-winner at HackUDC 2026 (1st Best Open Source AI, 2nd Inditex-Tech). Multi-layer visual search pipeline with YOLOv8, SigLIP-384 & DINOv2 ensemble over 27k+ fashion products, served via FAISS + FastAPI.

Project Image

LocalFold — Impacthon 2026 🥈

2nd Prize at GDG Challenge (Impacthon 2026). Web portal for protein structure prediction via AlphaFold2 without HPC expertise, featuring interactive 3D visualization (Mol*), PAE heatmaps and pLDDT confidence coloring. Multi-agent A2A system with 4 specialist agents powered by Gemini 2.5 Flash, deployed on Google Cloud Run.

Project Image

SmartPorts Ship Classification

Port CCTV image classification for ship detection and docking status using CNN and ResNet18 with stratified K-Fold cross-validation. Trained on real port imagery datasets.

Project Image

MatrixLens

Cache-aware audit tool for matrix multiplication kernels. Simulates L1 cache behavior (LRU, set-associative) to visualize how tiling strategies eliminate cache misses in high-performance computing.

Project Image

Check out more of my work on GitHub

GitHub Profile

Awards

On March 1, 2026, we presented our project at HackUDC for the InditexTech challenge. Following a stellar presentation before the Inditex team and GPUL, we were honored with the 1st Prize for Best Use of Open Source AI and the 2nd Prize in the Inditex-Tech Challenge.

🥇 1st Prize — Best Open Source AI
🥈 2nd Prize — Inditex-Tech Challenge

In April 2026, we competed at the GDG Challenge (Impacthon 2026) with LocalFold — a web platform for protein structure prediction using AlphaFold2, a multi-agent A2A system powered by Gemini 2.5 Flash, and deployment on Google Cloud Run. We were awarded 2nd Prize.

🥈 2nd Prize — GDG Challenge (Impacthon 2026)

Contact

Interested in collaborating or just want to say hi?

Say Hello